作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程

• 移动互联与通信技术 • 上一篇    下一篇

基于频域LMS算法的稀疏信道估计

傅剑斌,彭 华,董 政   

  1. (解放军信息工程大学信息系统工程学院,郑州 450002)
  • 收稿日期:2012-10-29 出版日期:2013-12-15 发布日期:2013-12-13
  • 作者简介:傅剑斌(1988-),男,硕士,主研方向:通信信号处理;彭 华,教授;董 政,博士

Sparse Channel Estimation Based on Frequency Domain Least Mean Square Algorithm

FU Jian-bin, PENG Hua, DONG Zheng   

  1. (Institute of Information System Engineering, PLA Information Engineering University, Zhengzhou 450002, China)
  • Received:2012-10-29 Online:2013-12-15 Published:2013-12-13

摘要: 传统自适应滤波方法无法直接、有效地对稀疏信道进行估计。为此,提出一种基于频域的稀疏信道估计方法。为削弱或消除信道的稀疏性质在其估计过程中带来的影响,引入频域最小均方(LMS)算法。频域LMS算法通过FFT变换实现稀疏信道的非稀疏化,从而使其可以对稀疏信道直接估计。仿真实验结果表明,频域LMS算法具有较好的收敛性能,与频域RLS算法相比,其收敛速度相当,但其收敛后的均方误差提高近10 dB,可较好地完成对稀疏信道的估计,同时在算法的实现过程中通过使用重叠保留法能较大程度地减少估计的运算量。

关键词: 自适应滤波, 稀疏信道, 频域信道估计, 频域最小均方算法, 重叠保留法

Abstract: As the traditional adaptive filtering algorithm can not effectively estimate the channel directly, a new frequency domain algorithm is proposed. With Frequency Domain Least Mean Squares(FD-LMS), it weakens the effect of the sparse character of the channel. It can change the sparse channel into nonsparsity by Fast Fourier Transformation(FFT), and the adaptive filtering algorithm estimates the channel directly. Simulation experimental results show that FD-LMS possesses excellent character of convergence, its convergence rate is almost equal to frequency domain RLS algorithm, however, its Mean Square Error(MSE) improvesby nearly 10 dB. Therefore, FD-LMS can estimate the sparse channel well, and the calculation for estimating the channel can also be reduced with overlap-save method.

Key words: adaptive filtering, sparse channe, frequency domain channel estimation, Frequency Domain-Least Mean Squares(FD-LMS) algorithm, overlap-save method

中图分类号: